KENNESAW, Ga. |
Sep 10, 2025

Artificial intelligence (AI) is often linked to supercomputers and massive data centers,
but Kennesaw State University researcher Bobin Deng is aiming for something a bit
more accessible through a new National Science Foundation (NSF) grant.  

An assistant professor in Kennesaw State’s College of Computing and Software Engineering, Deng said the goal is to move AI beyond the cloud and into the hands of people where it can have the most impact – their personal devices. The research could allow AI tools to function without an internet connection, something that is uncommon with many current systems. 

“Most of today’s AI runs on large, expensive servers,” Deng said. “But many applications don’t need that level of computing power. “If we can run AI directly on a mobile phone, a drone, or other smaller systems, then we remove the reliance on costly supercomputers and make AI easier and faster access in more situations.” 

At the heart of his project is a technique that uses activation sparsity, a process
where only a portion of neurons in an AI model are active while most remain inactive.
Instead of loading all the data into memory, the system accurately predicts in advance
which pieces it may need and loads only those, reducing memory usage, speeding up
processes, and lowering energy demands. 

“Other methods shrink models by reducing the precision of data or pruning less important parameters,” Deng said. “Our approach is different. We predict which values will be activated. That allows us to combine our method with existing techniques like pruning or quantization to make models even more efficient.” 

The project will test tiny machine– learning models as predictors to support larger systems. This balance makes AI more
feasible for devices like smartphones, drones, and industrial sensors. 

The open-source simulator Deng’s team is building will allow other researchers and students to explore and refine the technology.  

“This grant supports fundamental research that can benefit students in our new Master of Science in Artificial Intelligence program, as well as industry partners,” he said. “From monitoring factory robots to predicting equipment failures, embedding AI in smaller devices has the potential to transform many industries.” 

Deng conducts his research in the Sustainable Smart System Lab on KSU’s Marietta Campus. He discussed how the university has played a vital role in supporting his work, from lab space to grant administration.  

“The leadership here has been very supportive,” he said. “Whenever I have needed resources or assistance, from graduate assistants to proposal submissions, the university has been there to help.” 

CCSE Interim Dean Yiming Ji commended Deng’s efforts as a reflection of the college’s mission to advance innovation and prepare students for the future of technology.  

“Dr. Deng’s research represents the kind of forward-looking work that positions KSU as a leader in AI,” Ji said. “His approach not only addresses technical challenges but also creates real–world opportunities for students, industry partners, and the community.” 

– Story by Raynard Churchwell

Photos by Darnell Wilburn

Related Stories

A leader in innovative teaching and learning, Kennesaw State University offers undergraduate, graduate, and doctoral degrees to its more than 47,000 students. Kennesaw State is a member of the University System of Georgia with 11 academic colleges. The university’s vibrant campus culture, diverse population, strong global ties, and entrepreneurial spirit draw students from throughout the country and the world. Kennesaw State is a Carnegie-designated doctoral research institution (R2), placing it among an elite group of only 8 percent of U.S. colleges and universities with an R1 or R2 status. For more information, visit kennesaw.edu.

Dr. Geoffrey Hinton is a British-Canadian cognitive psychologist and computer scientist who won the Nobel Prize in Physics last year “for foundational discoveries and inventions that enable machine learning with artificial neural networks.” Between you and me, I got hosed. I should have been a winner.

The Nobel committee obviously overlooked my own entry entitled, “One molecule of glucose bound to one molecule of fructose will make sugar and winning the Nobel Prize sure would be sweet.” I don’t think they know a lot about physics over there in Norway.

Just as I am known as a modest yet much-beloved columnist who bears an uncanny resemblance to a young Brad Pitt, Dr. Hinton who looks nothing like Brad Pitt, young or old, is considered the Godfather of Artificial Intelligence. That’s like being Godfather of the Mafia. Only worse.

If somebody in the Mafia got out of hand, you would just shoot them or put them in a tub of concrete and deposit them in the East River. According to Dr. Hinton, artificial intelligence is likely to get rid of anybody left in the Mafia and the rest of us as well and it won’t need a gun or a sack of concrete to do it.

“It’s not inconceivable,” he has stated, “that artificial intelligence could wipe out humanity,” saying that there was a “10 to 20 percent chance” that AI would be the cause of human extinction within the following three decades. In fact, many experts expect AI to advance, probably in the next 20 years, to be “smarter than people.”

Admittedly, I am not the go-to person on the subjects of cognitive psychology and computer science (although knowing how sugar is made is pretty impressive), but I would posit that it is not going to take 20 years for artificial intelligence to get smarter than people. That’s already occurred in some instances. Just look at Congress.

Can you see a computer saying, “Beep! Beep! Hey, I want to suck up to Donald Trump. I think I will propose changing the name of Greenland to Red, White and Blueland and then he will get me elected to the Senate where I can do other dumb stuff. Boop!” There are some things a computer won’t do, even if a member of Congress will.

Dr. Hinton also worries about the impact of AI on religion. He says, “I think religion will be in trouble if we create other beings. Once we start creating beings that can think for themselves and do things for themselves, maybe even have bodies if they’re robots, we may start realizing we’re less special than we thought. And the idea that we’re very special and we were made in the image of God, that idea may go out the window.” An interesting observation.

Theologically speaking, if computers become robots, will there be girl robots and boy robots? If so, will boy robots let girl robots in the pulpit? Or will the boy robots tell other robots that if they think girl robots should be allowed to preach, they will be condemned to spend eternity in an electronic waste disposal bin at Best Buys?

As to whether or not we are made in the image of God, I believe that’s God’s call, not mine. Creation is His thing. I will say that had God asked me, there are a few people He created that I think we could just as soon done without. I couldn’t find His image in them with a flashlight. Maybe He just put them here to show us He has a sense of humor.

I probably won’t be around to see how all this plays out but despite the Godfather of AI’s ominous warning, no robot will ever make me feel less special. I’ve got a family that loves me more than I deserve. I have friends that have stood with me through the good times and the bad. I had a rewarding career. I am blessed to live in this special state in this special country.

Most of all, thanks to a benevolent editor willing overlook misplaced commas and grammatical errors (Is it who or whom?), I have the opportunity to share my thoughts with you each week and to receive your feedback. That may come in the form of a kudo or a rap on the knuckles. I suspect robots won’t give a flying algorithm for you or your opinions. I do. And there is nothing artificial about that.

You can reach Dick Yarbrough at dick@dickyarbrough.com or at P.O. Box 725373, Atlanta, Georgia 31139.

In a move that could reshape drug discovery, researchers at Harvard Medical School have designed an artificial intelligence model capable of identifying treatments that reverse disease states in cells.

Unlike traditional approaches that typically test one protein target or drug at a time in hopes of identifying an effective treatment, the new model, called PDGrapher and available for free, focuses on multiple drivers of disease and identifies the genes most likely to revert diseased cells back to healthy function.

The tool also identifies the best single or combined targets for treatments that correct the disease process. The work, described Sept. 9 in Nature Biomedical Engineering, was supported in part by federal funding.

By zeroing in on the targets most likely to reverse disease, the new approach could speed up drug discovery and design and unlock therapies for conditions that have long eluded traditional methods, the researchers noted.

Traditional drug discovery resembles tasting hundreds of prepared dishes to find one that happens to taste perfect. PDGrapher works like a master chef who understands what they want the dish to be and exactly how to combine ingredients to achieve the desired flavor.”

Marinka Zitnik, study senior author, associate professor of biomedical informatics in the Blavatnik Institute at HMS

The traditional drug-discovery approach – which focuses on activating or inhibiting a single protein – has succeeded with treatments such as kinase inhibitors, drugs that block certain proteins used by cancer cells to grow and divide. However, Zitnik noted, this discovery paradigm can fall short when diseases are fueled by the interplay of multiple signaling pathways and genes. For example, many breakthrough drugs discovered in recent decades – think immune checkpoint inhibitors and CAR T-cell therapies – work by targeting disease processes in cells.

The approach enabled by PDGrapher, Zitnik said, looks at the bigger picture to find compounds that can actually reverse signs of disease in cells, even if scientists don’t yet know exactly which molecules those compounds may be acting on.

How PDGrapher works: Mapping complex linkages and effects

PDGrapher is a type of artificial intelligence tool called a graph neural network. This tool doesn’t just look at individual data points but at the connections that exist between these data points and the effects they have on one another. 

In the context of biology and drug discovery, this approach is used to map the relationship between various genes, proteins, and signaling pathways inside cells and predict the best combination of therapies that would correct the underlying dysfunction of a cell to restore healthy cell behavior. Instead of exhaustively testing compounds from large drug databases, the new model focuses on drug combinations that are most likely to reverse disease.

PDGrapher points to parts of the cell that might be driving disease. Next, it simulates what happens if these cellular parts were turned off or dialed down. The AI model then offers an answer as to whether a diseased cell would happen if certain targets were “hit.”

“Instead of testing every possible recipe, PDGrapher asks: ‘Which mix of ingredients will turn this bland or overly salty dish into a perfectly balanced meal?'” Zitnik said.

Advantages of the new model

The researchers trained the tool on a dataset of diseased cells before and after treatment so that it could figure out which genes to target to shift cells from a diseased state to a healthy one.

Next, they tested it on 19 datasets spanning 11 types of cancer, using both genetic and drug-based experiments, asking the tool to predict various treatment options for cell samples it had not seen before and for cancer types it had not encountered.

The tool accurately predicted drug targets already known to work but that were deliberately excluded during training to ensure the model did not simply recall the right answers. It also identified additional candidates supported by emerging evidence. The model also highlighted KDR (VEGFR2) as a target for non-small cell lung cancer, aligning with clinical evidence. It also identified TOP2A – an enzyme already targeted by approved chemotherapies – as a treatment target in certain tumors, adding to evidence from recent preclinical studies that TOP2A inhibition may be used to curb the spread of metastases in non-small cell lung cancer.

The model showed superior accuracy and efficiency, compared with other similar tools. In previously unseen datasets, it ranked the correct therapeutic targets up to 35 percent higher than other models did and delivered results up to 25 times faster than comparable AI approaches.

What this AI advance spells for the future of medicine

The new approach could optimize the way new drugs are designed, the researchers said. This is because instead of trying to predict how every possible change would affect a cell and then looking for a useful drug, PDGrapher right away seeks which specific targets can reverse a disease trait. This makes it faster to test ideas and lets researchers focus on fewer promising targets.

This tool could be especially useful for complex diseases fueled by multiple pathways, such as cancer, in which tumors can outsmart drugs that hit just one target. Because PDGrapher identifies multiple targets involved in a disease, it could help circumvent this problem.

Additionally, the researchers said that after careful testing to validate the model, it could one day be used to analyze a patient’s cellular profile and help design individualized treatment combinations.

Finally, because PDGrapher identifies cause-effect biological drivers of disease, it could help researchers understand why certain drug combinations work – offering new biological insights that could propel biomedical discovery even further.

The team is currently using this model to tackle brain diseases such as Parkinson’s and Alzheimer’s, looking at how cells behave in disease and spotting genes that could help restore them to health. The researchers are also collaborating with colleagues at the Center for XDP at Massachusetts General Hospital to identify new drug targets and map which genes or pairs of genes could be affected by treatments for X-linked Dystonia-Parkinsonism, a rare inherited neurodegenerative disorder.

“Our ultimate goal is to create a clear road map of possible ways to reverse disease at the cellular level,” Zitnik said.